1903. J MARIKE FAUNA OF ZANZIBAR. 143 



there comes first a region of about forty segments bearing small 

 gills of three or fom- filaments ; these gills at about the sixtieth 

 segment rapidly enlarge and consist of four or fiv'e filaments, 

 though they never become long enough to meet over the back. 



So far as Grube's and Baird's desciiptions go, the last two species 

 correspond with M. janaarii and M. j^dTishi. Indeed the comb- 

 seta? of the latter form a distinct point of similarity between it 

 and AI. furcellata. But the descriptions published of these two 

 forms are so exceedingly meagre that certainty is impossible, so 

 that in preference to the risk of giving two worms a distribution 

 from East Tropical America to East Africa, I have described my 

 forms de novo. 



Genus Lysidice. 



Lysidice collaris Ehr., Grube. 



This species, though never abundant, occurred at most of my 

 collecting-places, viz., two specimens from Wasin Harbour (one 

 from between tide-marks, the other from a depth of 10 fathoms), 

 two smaller ones from 3 fathoms in Ohuaka Bay, and three, 

 smaller still, from the shore near Zanzibar Town. 



Discrepancies of some importance occur between the descriptions 

 of the species already published by Grube (Red Sea and Philippine 

 collections), Marenzeller (Japan), and Gravier (Red Sea). My 

 own collection shows that variations of features, usually considered 

 diagnostic, occur in specimens from the same locality. 



The name ' collaris ' obviously refers to the white ring found 

 near the anterior end of the living animal. As the colour dis- 

 appears from specimens which have been a few years in spirit 

 (it is already becoming faint in my own after the lapse of one 

 year), it has not yet been described. The ground-colour is a blight, 

 yellow-brown, best develojaed anteriorly and gradually dying out 

 at about the tenth setigerous segment. Posteriorly the body is 

 nearly colourless, unless sexual products^ which are pink, give it 

 that colour. In one of my specimens (and presumably in that 

 collected by Ehrenberg) this pigmentation is interrupted by 

 numerous small white dots, and is omitted altogether from 

 segments three and four, forming the above-mentioned white 

 collar. In the remaining specimens, one of which is of equal size 

 to this, the coloration is perfectly uniform. 



The form of the body is in life, as after preservation, flat below 

 and strongly arched above throughout its length. 



The insertion of the tentacles is not noticed by former authors 

 except Gravier, and in this respect, as in others, none of my 

 specimens agTee with his descrijDtion. The tentacles, though a 

 little narrowed at their bases, have no distinct basal joint, and 

 though they in some cases arise from nearly the same level, yet 

 the origin of the middle one is aUoays in front of the origin of the 

 other two, thus reversing the usual arrangement. The prostomium 

 itself is rather longer than in Gravier's fisui-e. 



